Trajectory prediction model of diabetes distress in adults with type 2 diabetes mellitus: a 12-month prospective longitudinal study.

INTRODUCTION

Clarifying heterogeneous diabetes distress (DD) trajectories and their predictors from a dynamic perspective is crucial. We aimed to develop a trajectory prediction model for dynamic DD.

METHODS

This prospective longitudinal study included 443 adults with type 2 diabetes mellitus who completed the demographics and diabetes characteristics questionnaire, scales measuring lifestyles and psychological factors (at baseline), and the Chinese version of the Diabetes Distress Scale (at baseline and at 3-, 6-, 9-, and 12-month follow-ups). After identifying the factors associated with DD, growth mixture modeling was used to determine latent longitudinal DD trajectory classes and develop a trajectory prediction model.

RESULTS

Five DD trajectories were identified: persistently low DD (65.01%), persistently moderate DD (25.28%), persistently high DD (3.61%), decreasing DD (3.16%), and increasing DD (2.94%). Using the persistently low DD group as the reference, people with no religious belief (B = -24.932, p < 0.001), longer diabetes duration (B = 0.042, p = 0.037), worse self-management behaviors (B = -0.032, p = 0.009), and lower self-efficacy (B = -0.287, p = 0.007) tended to have a persistently moderate DD trajectory. Insomnia severity (B = 0.232, p = 0.008) and type D personality (B = 2.783, p = 0.002) were significant positive predictors of persistently high DD trajectory. Those with higher HbA1c levels (B = 0.728, p = 0.003) and lower self-efficacy (B = -0.858, p = 0.044) were more likely to belong to the decreasing DD trajectory class. Self-management behaviors (B = -0.127, p = 0.012) were negatively associated with belonging to the increasing DD trajectory class.

CONCLUSION

Demographics, diabetes characteristics, lifestyles, and psychosocial factors can predict dynamic heterogeneous trajectories of DD. The trajectory prediction model will enable healthcare professionals to anticipate DD trajectories and conduct targeted interventions [Trial registration: ChiCTR2100047071].